outline

CODING Outline

The Method

The used method of the programs CODINGx, applying for a patent under the patent number DE 10 2005 033 390.7 A1 at the German patent office, is suitable for en-/decoding any data on place. The method is characterised by

the feature, that input data and output data (coded data) are of the same length,
a high operation speed (e.g. 2-digit MB value / sec. having a block length of 512 bytes , dimension t=1 and actual computer system > 2GHz),
blocks of data, independently processed, of (nearly) any block length (e.g. if a block can’t be decoded, all other blocks can; the block length can vary inside one dataset),
a very high safety, growing in any region by appropriately dimensioning the method (dimension t)
(it exists special documents

describing the method,
describing convergence of bits and bytes/groups of bytes against the rectangular distribution,
describing general safety of the used partial methods and the used method combinations,
describing the power of the key spaces and the state spaces of dimension t,
for the statistical analyses of near-key behaviour, of convergence behaviour, of block differences, etc. [at the moment especially of dimension t=1]).
With that CODING is the ideal tool

coding Voice over IP, blue tooth/wireless data,
coding multimedia information (e.g. MP3 data, DVD movies etc.),
coding any exchange of data (e.g. between customer/client and supplier in industry).

Performance

The time-critical coding of mass data is a special challenge. This among other things arises in a real-time scenario, in which a continuous data stream has to be handled in real time. Any delay by data processing being too slow causes a collapse and therefore cannot be used, i.e. it results in suspending the coding process.
The method CODING reduces the amount of multiplications being necessary to a minimum being required for maintenance of complete safety even if data is of most homogeneous type (tested case: 1 GB binary zeroes). Because of this the processing speed being significant faster in comparison with other methods will be achieved, where the CODING-method being more flexible than other methods simultaneously gives a higher safety (especially for t>1) on the other hand.

Operation Safety

Conventional coding systems as DES, Triple-DES, AES at fist are working with fixed block length, so shorter data first has to be expanded to such fixed length. At second blocks of such systems are entangled, i.e. a defective block (e.g. by transmission error) makes the decoding of all following blocks impossible.
CODING however has no fixed block length. I.e. shorter information at the end of a dataset is not expanded, so the length of the coded dataset is unchanged referring to the decoded data.

For the operation safety more essential is the fact that a damaged block don’t stop the decoding of the following blocks. Especially this is relevant for telephone conversations (Voice over IP), in which a trouble in data stream hasn’t have the automatic consequence to put up the call completely new. Also transmitting other data with CODING, it is only necessary to send the incorrect block once again. By modifying the here presented algorithm (e.g. retaining the restart point parameters of a defective block), the incorrect blocks can be processed later in a very fast and efficient way.

Encryption Safety

Even newer coding systems as AES use only a key with maximum length of 128 bits. In comparison already the simplest version of CODING (dimension t=1) uses a key with maximum length of 2048 bits (for dimension t=2 it already comprises 1,048,576 bits of maximum key length).
The method was submitted to information scientists of the TU Darmstadt conducted by Prof. Dr. Johannes Buchmann to get a first judgement. According to that the method can’t be broken without trying out the most part of all possible keys one after the other. An exact judgement of the safety especially in the view of information science (i.e. do you always need to check all possible keys for breaking a nontrivial key or can you reduce the number of possible keys to be checked) has to be left to a detailed examination.
Not depend on this, the author has made extensive tests, which prove that even if data is of most homogeneous type (only binary zeroes) the method converges against the rectangular distribution of binary zeroes/ones and bytes or byte groups respectively (special document see above).
Further near key judgements could show that even experimental keys differing in only one or few bits form original key don’t result in statistical relevant changes of distributions in relation to the coding data. Also inspecting block differences and its distributions couldn’t discover relevant differences to block contents by chance.
Rather the method presented here can be said to be safety even if besides the coding algorithm both the original data and the encoded data are known. It is impossible to derive the original key from these information (except testing all possible keys; see examination still be due).

Future Safety

The method CODING covers a (infinite) number of concrete coding processes. The dimension t of a process within CODING is defined as the number of bytes combined to a byte-group. For t=1 single bytes are looked at and 1-byte transformation tables are used. For t=2 each 2 bytes are interpreted as a separate byte-group, for which 2-bytes transformation tables are used, etc.
The method CODING with regard to t is not limited, at the moment PC-/Unix-programs fit for use exist until t=3.

Power of Key Space

For t=1 approx. 3,2317*10⁶¹⁶, for t=2 approx. 6,74114*10315.652, for t=3 approx. 10^{121.210.686,2336} possible key space elements exist!
The last number is describable by a single 1 followed by zeroes, which occupy more than 33 volumes with 1000 pages, any of them with 3600 zeroes. It is easy to prove that at least (256^t)! elements of the key space can be really effective by the algorithms mentioned before (n!, speak ”n factorial” is defined as 1*2*3*...*n), simply the proof of the effectiveness of all elements of a key space is due (see above).
Under these circumstances t has to choose only ”big enough” to give nobody even with future conceivable (quantum) computers or computer farms any realistic possibility to decode with CODING encoded data without corresponding key or coding parameters in acceptable time. At the moment this should be the case for t=3.

Characteristics

CODING is an ”on place” method, i.e. no change of data length.
CODING, with its high operational speed, is suitable for real time processes
CODING is usable for (nearly) any block length.
CODING represents

a very high operation safety,
an extreme high encryption safety and
an excellent future safety.

Development Status

The described method CODING has its beginning in preliminary examinations of the author approximately 19 years ago. The consistent development, generalization, and implementation in Free Pascal programs up to now for the dimensions t=1, t=2, and t=3 recently have been driven on the basis of changed requirements on part of the potential users and the possibilities of computers on part of the implementation.

The CODING programs are completely tested and fit for use for t=1, t=2, and t=3.

back to the top